overheating cables

[dawsirn]

Was called to a property today where the night before they had smoke filling up a room and the smell of burning electrical cables. By the time I arrived later today (was only called an hour previous to my arrival) a builder who was working on a extension to the property had already cut out sections of the wall which had black charring to see if he could find the problem.

So when I arrived the customer explained the existing setup and I was surprised to see a previous extension the property with it's own submain was spurred of the ring. Now this spur wasn't even fused down. It was just a 45a DP switch supplying a 100a main switch consumer unit with various MCB's supplying various circuits ranging from sockets, lights and an outbuilding.

So this dodgy setup has been doing it's thing for a few years until it went up in smoke yesterday. After investigating a bit further where the builders had already started digging away I noticed one of the ring cables was installed behind the skirting board which is obviously bad enough but to make matters worse there's a wet under floor heating system throughout! all the cable I pulled back was showing signs of overheating. This was so bad it appears to have actually blackened the floor all along the bottom of the skirting board. All ceramic tiles. The builder had already pulled out celotex which was charred and the inside of the metal capping on the ring cable was charred also.

I think the straw the broke the camels back was the building work currently going on, as the tools etc are all plugged into sockets supplied from this spurred consumer unit.

In case anyone was wandering I decided to completely disconnect the circuit until further investigation and remedial work is carried out. I also found a Insulation resistance reading of 0.5Mohm between the Live and Earth on that circuit too. end to end readings were fine. r1 0.37 rn 0.37 r2 0.60


I've never seen it quite this bad before. I was wandering how hot the cable would have to be for it to actually cause this sort of damage?
Would the entire circuit be compromised now and full rewire of circuit recommended, or would I just have to try and replace local cables to where the heat damage was found. Not sure why the low insulation resistance reading exists but I've already recommend that is found too.

I've attached pictures but apologies for the crappy quality.

 

[Marvo]

To damage a cable along its entire length like that and fill the room with smoke it must have been a short circuit or pretty close to one. PVC won't start smoking until it reaches way over 100 degrees Celsius, probably nearer 150C in fact.

 

[Lucien Nunes]

What OCPD was feeding the RFC? If all the load up to the rating of a 32A MCB was concentrated near one end and went via one leg due to the bogus spur, then 2.5 could be overloaded but not by all that much, not enough to produce smoke. If the OCPD was incorrect, bypassed or faulty, the entire cable could have been damaged by gross overload.

 

[dawsirn]

What OCPD was feeding the RFC? If all the load up to the rating of a 32A MCB was concentrated near one end and went via one leg due to the bogus spur, then 2.5 could be overloaded but not by all that much, not enough to produce smoke. If the OCPD was incorrect, bypassed or faulty, the entire cable could have been damaged by gross overload.

The RFC was fed via a 32a MCB with additional protection from 30ma RCD. I should mention the customer reported that none of these tripped during all this.

So it's unlikely that just a overloaded cable alone would cause this much heat? even being installed so close to the underfloor heating? The customer had a photo of another room with all the exposed UFH pipes and the plumber had run the pipes right up against the edge of the floor so I suspect the same type of install in this room too.

I need to go back and try and narrow down where the low insulation resistance reading is to see if a short circuit plays any part in this.

thanks

 

[gutterball]

You need to lock that switch off not put tape across it.

 

[gnselectricals]

The RFC was fed via a 32a MCB with additional protection from 30ma RCD. I should mention the customer reported that none of these tripped during all this.

So it's unlikely that just a overloaded cable alone would cause this much heat? even being installed so close to the underfloor heating? The customer had a photo of another room with all the exposed UFH pipes and the plumber had run the pipes right up against the edge of the floor so I suspect the same type of install in this room too.

I need to go back and try and narrow down where the low insulation resistance reading is to see if a short circuit plays any part in this.

thanks

I would be worried about low IR values with cabling like that tbh would expect something has shorted together there at least.

 

[Deleted member 72666]

500,000Ω doesn't sound like a short to me, it's more likely to be the charring at an accessory or cable termination tracking to earth.

 

[dawsirn]

You need to lock that switch off not put tape across it.

The tape is just for information purposes. I've disconnected the RFC live and neutrals from main consumer unit.

Quick update, I went back to try and narrow down the insulation resistance fault and to have a more in depth look but on arrival customer informs me they've been having drops in voltage throughout the house. Checked main cut out and found high earth fault loop of 15ohms. It's a overhead PME system. Same reading between L-N. So have contacted Western Power who are treating it as an emergency and will hopefully be there soon.

Advised they keep the power off until there arrival.

What the customer failed to mention yesterday is that this issue has been intermittent all week so now I'm thinking this has had some bearing on the main issue of the overheating cable, or am I barking up the wrong tree?

Had to shoot of but will call in again to try and narrow down the fault and hear what Western Power had to do.

 

[telectrix]

keep us posted.

 

[Lucien Nunes]

Does it look like the CPC overheated rather than the L & N? Overload due to the incorrect feeding of the sub-board off the RFC would have mostly concerned the L & N, and as I said probably on a 32A wouldn't have burnt the cable under any circumstances. But if the loop impedance was too high on a PME then all sorts of overcurrent could have been flowing back down the CPC to an incidental earth somewhere on the circuit like a casual contact with an extraneous metal part of lower impedance than the faulty supply CNE. In fact the entire installation current might have been going down that CPC, regardless of whether the circuit was itself overloaded.

 

[Dave OCD]

Quite possibly a high resistance on the neutral connection at the WPD pole.

 

[DPG]

Any confirmation of exactly what this issue was? Daz

 

[dawsirn]

sorry, Haven't had exact confirmation on what wpd distribution found, They did leave a message on my phone saying they had to cut back some trees but will know more when I see the client Friday. Will try and arrange a full EICR of the property soon to try and find any issues that may relate to the initial problem. Circuit is still disconnected so need to try and get back in there and rectify the issue but completely stacked out with other commitments at the moment!

Will keep you posted with any findings as I may need to pick your brains when I have all the results.

thanks

 

[dawsirn]

Quick update on this one.

Called round to see client to discuss plans for rectifying this fault and other ongoing works planned to another part of property.

Anyways it turns out WPD found a bad neutral connection on one of their poles which supply the property. ( picture attached of the pole in question).

Only had a brief amount of time there again today but traced the insulation resistance fault to another part of the ring which is supplying another out building 50 meters down the garden (again not fused down etc but doesn't seem related to this issue, from what I can tell anyway)

So pulled out more of the overheated cable from under the skirting board and there's no real obvious point where the heat's being generated. It's darker in a few places nearer the socket with the dodgy spur. The further away from that point the cleaner the cable looks. It's testing out (IR and end to ends) fine so will probably cut out the visibly darker section and install new section of cable properly.

The dodgy spur will be removed and new supply cable from the consumer unit to submain will be installed instead.

So still not sure exactly why the cable has overheated like this, but I'm assuming it's not a coincidence that the bad neutral connection happened at the same time. The dodgy spur would have been running a large electric heater (haven't checked loading yet) and a few other things at the time of the overheating.

I'm still planning to carry out a few other checks on the property in due course so If anyone has any suggestions on what to look for I'd be very grateful as this is a fault I haven't come across before!

I've attached a few more photos of the affected area.

 

[Marvo]

I still don't see how a floating neutral could have caused that kind of damage to a cable. I think it was a short circuit or massive overload that didn't get disconnected in anywhere near the recommended time by the OCPD.

 

[Lucien Nunes]

It's a TN-C-S supply and we can infer from the loop impedance tests that it was not just a neutral but a CNE conductor that failed (15 ohms to both E & N). Depending on the non-electrical services and bonding, the entire installation load current could have been diverted at the supply terminals back through any CPCs that had an earth anywhere along their length, as presumably the transformer remained earthed at the pole. If the best earth was on something connected to this final, a 1.5 or even 1.0 CPC might have been carrying the full load including any showers etc. A short circuit on anything more than a 6A MCB would not have been cleared at all

I would be suspect thermal damage to any and all cables that have say <10 ohms to earth at any point beyind the CU and would test for this. Check whether the DNO will bear cost of repairs under the terms of supply in the ESQCR.

 

[Marvo]

As you say it's plausible with a poor CNE on a TN C S arrangement but there would need to be very few parallel earth paths within the property and lack of bonding or plastic services for water, gas etc to complete the perfect storm for a very dangerous situation. If the CPC was clearly the overheated conductor I'd also lodge a claim against the DNO.